There are many dApps and tools on Hive that are dependent on Hive public nodes. Anyx is one of the most used hive nodes and is also a very powerful node. Most of the developers use that node and when something goes wrong with this node it is easily noticeable. Yesterday a similar glitch happens and even today the nodes are not fully ready. There have been issues with the nodes. 

I usually get lots of error messages related to Hivemind in my logs. That's the usual thing. My applications are dependent on getting the post information as soon as a post/comment is published. I stream the blocks and then make API calls to get the content of the post/comment. If there is a delay in Hivemind the API would throw an error message. There are a few public nodes that support Hivemind and most of the developers are dependent on those public nodes as it is very costly to have our own Hivemind instance. 

Yesterday I suddenly started receiving lots of error messages in my logs. Upon investigation, I noticed that my application was pointing to the Anyx.io node. Surprisingly dHive has a failover built-in but I was wondering why it was not triggered when this error happened. There is a good tool on Peakd that provides the [status of all the popular Hive nodes](https://beacon.peakd.com/). We can use that to check the status of the nodes. 

![image.png](https://images.hive.blog/DQmYTQQE5FFZhR4ycPSoTKfUcY5ruXKgC1hjqJDUYdmLrMo/image.png)

## Need stable nodes

I have to admit that nodes have been super stable and a lot better than how it used to be on Steem. A big kudos to that but I personally feel that it is not going to be enough. Especially after HF24, we are able to see so many issues mostly related to Hivemind. There are delays at times when Hivemind and nodes go off sync. 

Though we have so many use cases for Hive, still blogging is one of the primary use case and I personally feel that we should be able to seamlessly retrieve the post/comment information and information about votes as soon as it made. We are able to stream that information from the blocks but if we need more information on that, we have to make a call to the hivemind to get the full content of the post. That is when we are having lots of problems right now. I'm not even sure how this is going to be fixed or if there is even a plan to fix this. 

![image.png](https://images.hive.blog/DQmRovZyNQRd6qi6xJ1ttkonZxBvvLk7iAr7hJ2ua81xFa6/image.png)

[Source](https://pixabay.com/photos/child-tower-building-blocks-blocks-1864718/)

Sometimes this is what makes me feel that we will need more stable nodes before we look for the price to pump or before doing any campaign on Twitter or other social platforms. We all know that Blogging is the only decentralized use case we have right now for Hive. All other gaming use cases are mostly centralized solutions or something that runs on a centralized side chain. If we are looking for more dApps and new projects, we would need stable nodes. I'm not talking about complete error-free nodes but at least we can reduce the outage. 

The big advantage that we have on Hive Blockchain is that even if nodes go offline or have glitches, we still have security where there is no data loss. The dApps are easily able to adapt to these glitches because of this very reason. But we cannot keep that as an excuse. I understand that we have a strong team led by @blocktrades already working on improving Hivemind. 

## Solution I'm excited about

In one of the previous articles, @blocktrades mentioned that they are looking for options to modularize Hivemind (suggested by @howo) so that the dApps can have their own instance of Hivemind with just one or few modules enabled that would benefit the dApp. I'm really looking forward to this solution. 

![image.png](https://images.hive.blog/DQmao74MJXkBK2k4zgcMR8WHWciZjTDAHG8htpShpiiyPuY/image.png)

[Source](https://pixabay.com/illustrations/building-blocks-play-game-blocks-615239/)

In my case, I'm interested to see only the posts/comments/votes that are published on the blockchain. If I can set up a Hivemind instance only for that and also have a facility where the instance would retain data only till payout which is 7 days, it would be an optimized solution for me. I hope this will make things better for the dApps that I try to run. Instead of being dependent on one of the public nodes, I would even consider having my own instance running. 

---

<center>[![Kani Bot notification service](https://files.peakd.com/file/peakd-hive/bala41288/DDxrJxFV-kani-banner-15fps.gif)](https://discord.com/invite/S4x6PFz)
[
![image.png](https://files.peakd.com/file/peakd-hive/bala41288/46eaz12N-image.png)](https://discord.gg/bGmS2tE)
[![image.png](https://images.hive.blog/DQmag62WZkMSrJAx5bupB6C8QSTdzQfZoCMsk5dp5kEjn1v/image.png)](https://r.honeygain.me/BALA4F8CD5)
[![image.png](https://files.peakd.com/file/peakd-hive/bala41288/9hTRE5gg-image.png)
](https://netbox.global/r/balanbx)

<a href="https://www.publish0x.com?a=YRdG6knybD"><img src="https://cdn.publish0x.com/prod/fs/images/da2424398bd4625ed05e95650fd1fc59638d98e68350cf673d8b393558969644.png" width="728" height="90" /></a>
</center>

---

Posted Using [LeoFinance <sup>Beta</sup>](https://leofinance.io/@bala41288/if-we-need-more-investors-and-dapps-we-will-need-more-stable-nodes)